RWA Microsoft’s AI head has raised concerns that Anthropic’s Claude model is becoming increasingly human-like, which could heighten the risk of losing control over advanced AI systems. The warning underscores growing industry worries about AI safety as models develop more human-like behaviors.